About AlphaDog
Your due diligence analyst for US equities. Every claim, sourced.
The agents do the research. You get the control room — live progress, fact-checked numbers, a full audit trail, and a sourced Buy/Hold/Avoid verdict in about 60 seconds.
The problem we kept seeing
Equity research is either gated behind five-figure terminals or buried in noisy newsletters that never look back at their own calls. Retail tools generate reports that read like marketing copy with no audit trail.
Meanwhile every honest framework — Buffett, Lynch, the forensic accounting literature — agrees: the value is in unglamorous, sourced, repeatable analysis. Not vibes.
What we do differently
Thirteen specialized agents — market, filings, financials, news, ownership, valuation, catalysts, debate, critic, contrarian — produce a report where every claim is sourced and every number is auditable.
Deterministic factor scores for quality, growth, value, safety, and momentum anchor the verdict before the narrative is written — and forensic accounting checks surface the red flags most retail reports quietly skip.
Core Philosophy
What we believe
No hype, ever
No price targets pulled from thin air, no breathless newsletter spin. Reports stick to what the filings, financials, and market data actually support.
Sourced or it didn't happen
Each bullet in every report links back to an SEC filing, a news article, or a market data point. No black-box reasoning, no unverifiable claims.
Numbers before narrative
Market data, financial statements, filings, valuation, and news are checked before the report turns them into a plain-English thesis.
You own the thesis
Reports surface thesis-breaker triggers — objective conditions that would invalidate the call. That's a research tool, not a tip sheet.
Methodology
How the pipeline produces a verdict
13-agent pipeline
Resolver → market → SEC filings → financials → news → ownership → scoring → valuation → catalysts → synthesis → critic → debate → contrarian.
Deterministic scoring
Factor scores for quality, growth, value, safety, and momentum anchor the verdict before the narrative is written.
Draft + verify
Synthesis drafts the narrative; the critic agent fact-checks every number and flags anything unverified.
How we test
Every model change runs against a golden set of tickers — agent success, citation count, verdict validity, and latency — before it ships.